Finer-Resolution Mapping of Global Land Cover: Recent Developments, Consistency Analysis, and ProspectsRead the full article
The Journal of Remote Sensing, an Open Access journal published in association with AIR-CAS, promotes the theory, science, and technology of remote sensing, as well as interdisciplinary research within earth and information science.
The Journal of Remote Sensing’s editorial board is led by Yirong Wu (Aerospace Information Research Institute, Chinese Academy of Sciences) and is comprised of experts who have made significant and well recognized contributions to the field.
Journal of Remote Sensing is now accepting submissions for its first two special issues:
Latest ArticlesMore articles
Sensitivity of Estimated Total Canopy SIF Emission to Remotely Sensed LAI and BRDF Products
Remote sensing of solar-induced chlorophyll fluorescence (SIF) provides new possibilities to estimate terrestrial gross primary production (GPP). To mitigate the angular and canopy structural effects on original SIF observed by sensors (SIFobs), it is recommended to derive total canopy SIF emission (SIFtotal) of leaves within a canopy using canopy interception () and reflectance of vegetation (). However, the effects of the uncertainties in and on the estimation of SIFtotal have not been well understood. Here, we evaluated such effects on the estimation of GPP using the Soil-Canopy-Observation of Photosynthesis and the Energy balance (SCOPE) model. The SCOPE simulations showed that the between GPP and SIFtotal was clearly higher than that between GPP and SIFobs and the differences in () tend to decrease with the increasing levels of uncertainties in and . The resultant decreased to zero when the uncertainty level in and was ~30% for red band SIF (RSIF, 683 nm) and ~20% for far-red band SIF (FRSIF, 740 nm). In addition, as compared to the TROPOspheric Monitoring Instrument (TROPOMI) SIFobs at both red and far-red bands, SIFtotal derived using any combination of (from MCD15, VNP15, and CGLS LAI products) and (from MCD34, MCD19, and VNP43 BRDF products) showed comparable improvements in estimating GPP. With this study, we suggest a way to advance our understanding in the estimation of a more physiological relevant SIF datasets (SIFtotal) using current satellite products.
Multisensor Remote Sensing Imagery Super-Resolution with Conditional GAN
Despite the promising performance on benchmark datasets that deep convolutional neural networks have exhibited in single image super-resolution (SISR), there are two underlying limitations to existing methods. First, current supervised learning-based SISR methods for remote sensing satellite imagery do not use paired real sensor data, instead operating on simulated high-resolution (HR) and low-resolution (LR) image-pairs (typically HR images with their bicubic-degraded LR counterparts), which often yield poor performance on real-world LR images. Second, SISR is an ill-posed problem, and the super-resolved image from discriminatively trained networks with norm loss is an average of the infinite possible HR images, thus, always has low perceptual quality. Though this issue can be mitigated by generative adversarial network (GAN), it is still hard to search in the whole solution-space and find the best solution. In this paper, we focus on real-world application and introduce a new multisensor dataset for real-world remote sensing satellite imagery super-resolution. In addition, we propose a novel conditional GAN scheme for SISR task which can further reduce the solution-space. Therefore, the super-resolved images have not only high fidelity, but high perceptual quality as well. Extensive experiments demonstrate that networks trained on the introduced dataset can obtain better performances than those trained on simulated data. Additionally, the proposed conditional GAN scheme can achieve better perceptual quality while obtaining comparable fidelity over the state-of-the-art methods.
Assessment of Slope-Adaptive Metrics of GEDI Waveforms for Estimations of Forest Aboveground Biomass over Mountainous Areas
Waveform broadening effects of large-footprint lidar caused by terrain slopes are still a great challenge limiting the estimation accuracy of forest aboveground biomass (AGB) over mountainous areas. Slope-adaptive metrics of waveforms were proposed in our previous studies. However, its validation was limited by the unavailability of enough reference data. This study made full validation of slope-adaptive metrics using data acquired by the Global Ecosystem Dynamics Investigation (GEDI) mission, meanwhile exploring GEDI waveforms on estimations of forest AGB. Three types of waveform metrics were employed, including slope-adaptive metrics (RHT), typical height metrics relative to ground peaks (RH), and waveform parameters (WP). In addition to terrain slopes, two other factors were also explored including the geolocation issue and signal start and ending points of waveforms. Results showed that footprint geolocations in the first version GEDI data products were shifted to the left forward of nominal geolocations with a distance of about 24 m~30 m and were substantially corrected in the second version; the fourth and fifth groups of signal start and ending points of waveforms had worse performance than the rest of the four groups because they used the maximum and minimum signal thresholds, respectively. Taking airborne laser scanner (ALS) data as reference, the root mean square error (RMSE) of terrain slopes extracted from the digital elevation model of the shuttle radar topography mission (SRTM DEM) was about 3°. The coefficients of determination () of estimation models of forest AGB based on RH metrics were improved from 0.48 to 0.68 with RMSE decreased from 19.7 Mg/ha to 15.4 Mg/ha by the second version geolocations. The RHT and WP metrics gave the best and the worst estimation accuracy, respectively. RHT further improved to 0.77 and decreased RMSE to 13.0 Mg/ha using terrain slopes extracted from SRTM DEM with a resolution of 1 arc second. of estimation models based on RHT was finally improved to 0.8 with RMSE decreased to 11.7 Mg/ha using exact terrain slopes from ALS data. This study demonstrated the great potential of slope-adaptive metrics of GEDI waveforms on estimations of forest aboveground biomass over mountainous areas.
Annual Maps of Forests in Australia from Analyses of Microwave and Optical Images with FAO Forest Definition
The Australian governmental agencies reported a total of 149 million ha forest in the Food and Agriculture Organization of the United Nations (FAO) in 2010, ranking sixth in the world, which is based on a forest definition with . Here, we report a new forest cover data product that used the FAO forest definition ( and at observation time or mature) and was derived from microwave (Phased Array type L-band Synthetic Aperture Radar, PALSAR) and optical (Moderate Resolution Imaging Spectroradiometer, MODIS) images and validated with very high spatial resolution images, Light Detection and Ranging (LiDAR) data from the Ice, Cloud, and land Elevation Satellite (ICESat), and in situ field survey sites. The new PALSAR/MODIS forest map estimates 32 million ha of forest in 2010 over Australia. PALSAR/MODIS forest map has an overall accuracy of ~95% based on the reference data derived from visual interpretation of very high spatial resolution images for forest and nonforest cover types. Compared with the canopy height and canopy coverage data derived from ICESat LiDAR strips, PALSAR/MODIS forest map has 73% of forest pixels meeting the FAO forest definition, much higher than the other four widely used forest maps (ranging from 36% to 52%). PALSAR/MODIS forest map also has a reasonable spatial consistency with the forest map from the National Vegetation Information System. This new annual map of forests in Australia could support cross-country comparison when using data from the FAO Forest Resource Assessment Reports.
Impact of Temperature on Absorption Coefficient of Pure Seawater in the Blue Wavelengths Inferred from Satellite and In Situ Measurements
There has been a long history of interest on how (if) the absorption coefficient of “pure” fresh water () and “pure” seawater () changes with temperature (), yet the impact of reported in the literature differs significantly in the blue domain. Unlike the previous studies based on laboratory measurements, we took an approach based on ~18 years (2002–2020) of MODIS ocean color and temperature measurements in the oligotrophic oceans, along with field measured chlorophyll concentration and phytoplankton absorption coefficient, to examine the relationship between and the total absorption coefficient () at 412 and 443 nm. We found that the values of and in the summer are nearly flat (slightly decreasing) for the observed range of ~19–27 °C. Since there are no detectable changes of chlorophyll during this period, the results suggest that has a negligible impact on and in this range. As a complement, the impact of salinity on was also evaluated using three independent determinations of and , where good agreements were found from these observations.
Feature Enhancement Network for Object Detection in Optical Remote Sensing Images
Automatic and robust object detection in remote sensing images is of vital significance in real-world applications such as land resource management and disaster rescue. However, poor performance arises when the state-of-the-art natural image detection algorithms are directly applied to remote sensing images, which largely results from the variations in object scale, aspect ratio, indistinguishable object appearances, and complex background scenario. In this paper, we propose a novel Feature Enhancement Network (FENet) for object detection in optical remote sensing images, which consists of a Dual Attention Feature Enhancement (DAFE) module and a Context Feature Enhancement (CFE) module. Specifically, the DAFE module is introduced to highlight the network to focus on the distinctive features of the objects of interest and suppress useless ones by jointly recalibrating the spatial and channel feature responses. The CFE module is designed to capture global context cues and selectively strengthen class-aware features by leveraging image-level contextual information that indicates the presence or absence of the object classes. To this end, we employ a context encoding loss to regularize the model training which promotes the object detector to understand the scene better and narrows the probable object categories in prediction. We achieve our proposed FENet by unifying DAFE and CFE into the framework of Faster R-CNN. In the experiments, we evaluate our proposed method on two large-scale remote sensing image object detection datasets including DIOR and DOTA and demonstrate its effectiveness compared with the baseline methods.